Re: [pcp] qa/667 (pcp2graphite) failing after latest merge

[Ken McDonell <kenj@xxxxxxxxxxxxxxxx>]

On 29/07/15 08:48, Ken McDonell wrote:
> On 29/07/15 07:17, Ken McDonell wrote:
> ...
> But I have another problem ... the "od" part of 667 is not working and
> there are lots of differences ... we need a smarter check here.
>
> Current 667.out.bad attached.
Digging a little deeper ... has the pcp2graphite pickle format 0 changed 
is pcp2graphite generating a different format?
We used to have in the output

V(metricname)\n

and we now have

S('metricname')\n

that's not all of the differences, but it sure is a lot of 'em.

Given that this is mangled by od(1), a remake of 667.out is the only 
option I have, unless someone can verify that the 667.out.bad I posted 
is now the expected output of qa/667.